This component of our CREATE proposal uses fenced funding from the Mental Health Strategic Plan to investigate the effects of evidence-based treatment on health outcomes for individual Veterans. In addition, this study will evaluate the effectiveness of the fenced funding approach for improving access to and quality of specialty SUD treatment services. Our findings will help the Office of Mental Health Services refine its efforts to measure and promote access and quality in the future and inform current discussions regarding whether fenced funding should be used in the coming years. Our operational partners in this research will be the Office of Mental Health Operations, the Program Evaluation and Resource Center, and the Office of Mental Health Services. The findings of this study will also inform VA clinicians and managers at the station and VISN levels about the effects of evidence-based treatment on health outcomes for individual Veterans. Background As part of a mental health strategic plan initiative, between 2005 and 2008 the VA provided targeted funding for specialty substance use disorder (SUD) treatment programs. This SUD fenced funding was intended to improve access to SUD treatment. In total, about $150 million were allocated for SUD treatment as part of the larger initiative. This project complements the other projects in this CREATE proposal because it applies some of the same performance metrics featured in the other projects to important policy questions and it investigates the mediating effects of these variables on outcomes for Veterans. Objectives Aim 1: Determine the proportion of fenced funds that can be accounted for by expanded staffing for specialty substance abuse treatment services. Aim 2: Estimate the effect of fenced funding-related staffing on access and process quality. We will use 18 access and process quality metrics developed and currently used by OMHO. Aim 3: Measure the effects of fenced funding-related staffing on patient-level outcomes. Patient-level outcomes will include hospitalization for a SUD-related condition, readmission to inpatient or residential treatment, and incident homelessness. Methods This will be a retrospective, observational analysis of administrative data. The principal population studied will be a cohort of approximately 500,000 Veterans who were diagnosed or treated for a substance use disorder in 2003 or 2004. We will follow these patients through 2011 to assess patient outcomes under Aim 3. Analyses under Aims 1 and 2 will examine the population of VA facilities from 2004 through 2011. Statistical models will use station-level fenced funding allocations as instrumental variables to isolate quasi-random variations in SUD staffing leading to changes in access and quality metrics and patient-level outcomes.